HIV-1 is not cured despite therapy that reduces viral replication to nearly undetectable levels, because reservoirs of virus persist in the host. These reservoirs persist because the virus establishes latent infection, primarily in CD4-positive memory T cells. These cells reactivate at a low but relatively constant frequency, leading to recrudescent viremia within 1-2 weeks after drug-therapy is stopped. These reactivation events are presumably stochastic, but we hypothesize that they can be induced by the specific antigens to which the cells are programmed to respond. Here, we propose to test this hypothesis and define those antigens, toward the goal of developing new approaches to the reactivation of latently infected cells that are potent yet selective. During the R21 phase, Specific Aim 1 will determine whether latency can be reversed in cells from infected patients by stimulation with HIV-1 antigens. HIV-infection favors cells that respond immunologically to the virus itself, so we hypothesize that many if not most latently infected CD4-positive memory T cells will respond to HIV antigens. To test this, we will measure the reactivation of viral gene expression ex vivo in latently infected cells from HIV-infected persons, first using peptide pools corresponding to HIV-1 proteins, and second, using replication-defective but authentic whole virions, which we can produce in large amounts, as antigens. By comparing the reactivation by HIV-1 antigens to that of maximal T cell stimulation, we will determine how effective specific antigenic stimulation might be. Specific Aim 2 will determine whether latency can be reversed in infected CD4-positive T cells by antigens other than HIV-1 to which the patient likely has immunologic memory. Specific Aim 3 will compare the abilities of these antigens to reverse latency with their abilities to stimulate the overall CD4 T cell population; this tests the hypotheis that latency is differentially established in cells with different antigen-specificities, and in particular that it is established preferentially in HIV-specific cells. The identification of antigns, either from HIV or other pathogens, that activate a substantial fraction of latently infected cells from patients ex vivo is the milestone for progression to the R33 phase. The R33 phase will provide in depth study of the potential of the antigens tested above to be used in therapeutic vaccinations, toward the goal of immunologically reactivating and eliminating latently infected cells. Specific Aim 4 will determine the broad applicability of the approach by optimizing antigenic mixtures and applying them to a larger cohort of patient cells. Specific Aim 5 will determine whether latently infected cells reactivated by specific antigens die spontaneously or can be killed by various aspects of innate and acquired cytotoxic immunity. These studies should facilitate the design of clinical trials, the goal of which would be to reactivate latently infected cells by provision of antigens to which the cells are programmed to respond, and to promote cytotoxic immunity that recognizes and kills those cells.